For me it was about making a game from start to finish, and adding in a new feature every time. So it would have a start screen, basic menu, a final level, a win screen and a way to restart. The first few games I wrote did not have a save, but they were short enough that someone could play it through in a single session. They had at least 10 levels, and it takes about 2-3 minutes to complete a level and two to three difficulty settings so that there was replay ability. Every game was a little longer, had more options etc.
With an adventure game - work on how many scenes, or how many interactions, so consider each scene a level, and how long the player is going to walk around for i.e. how big the scene is and how much dialogue there is. It's also something you learn with experience - you need to first create a playable scene and then time yourself - if it's taking you 10 seconds or 10 minutes, there's a problem. (the first is too short - put more obstacles in the way, the latter is too long - players will get bored, have more scene changes.)
My criteria for free games is that it should not take someone longer to find, download and install the game than to play it. And it should be a game they can replay to make it worth their while i.e. it will be different next time because of random elements. (That's a little harder to do in adventure games.)
My personal criteria for the the paid games I develop is that it should be at least 5 hours of game play (which includes reading the stories - those that have them in) - to justify the price. (That's more or less the benchmark for casual games - I read a lot of forums / reviews to get a feel for what people liked / disliked - and below that benchmark people often ask for their money back - longer than that adds more value, you can charge more or they feel that they are getting good value for money - this is at the $2.99 - $6.99 price point.)
You can find good quality adventure games here (free) that will give you an idea.